
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import math
import warnings

import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split, cross_val_score
from sklearn.preprocessing import OneHotEncoder, StandardScaler
from sklearn.compose import ColumnTransformer
from sklearn.pipeline import Pipeline
from sklearn.metrics import precision_score, recall_score, accuracy_score, f1_score
from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier, AdaBoostClassifier
from sklearn.svm import SVC
from sklearn.neighbors import KNeighborsClassifier
from sklearn.naive_bayes import GaussianNB, BernoulliNB
from sklearn.neural_network import MLPClassifier
from sklearn.linear_model import SGDClassifier
import xgboost as xgb
import lightgbm as lgb
from catboost import CatBoostClassifier

/Users/kapilwanaskar/anaconda3/envs/HW3/lib/python3.8/site-packages/scipy/__init__.py:146: UserWarning: A NumPy version >=1.16.5 and <1.23.0 is required for this version of SciPy (detected version 1.24.4
  warnings.warn(f"A NumPy version >={np_minversion} and <{np_maxversion}"


df = pd.read_csv('training_wids2024C1.csv')

#shape of the data
print("shape of the data is: ", df.shape)

df.head()

shape of the data is:  (12906, 83)


# Increase display option for maximum rows to ensure all columns are shown
pd.set_option('display.max_rows', 100)

# print list of all 83 columns and null values in each column
print(df.isnull().sum())

patient_id                                   0
patient_race                              6385
payer_type                                1803
patient_state                               51
patient_zip3                                 0
patient_age                                  0
patient_gender                               0
bmi                                       8965
breast_cancer_diagnosis_code                 0
breast_cancer_diagnosis_desc                 0
metastatic_cancer_diagnosis_code             0
metastatic_first_novel_treatment         12882
metastatic_first_novel_treatment_type    12882
Region                                      52
Division                                    52
population                                   1
density                                      1
age_median                                   1
age_under_10                                 1
age_10_to_19                                 1
age_20s                                      1
age_30s                                      1
age_40s                                      1
age_50s                                      1
age_60s                                      1
age_70s                                      1
age_over_80                                  1
male                                         1
female                                       1
married                                      1
divorced                                     1
never_married                                1
widowed                                      1
family_size                                  4
family_dual_income                           4
income_household_median                      4
income_household_under_5                     4
income_household_5_to_10                     4
income_household_10_to_15                    4
income_household_15_to_20                    4
income_household_20_to_25                    4
income_household_25_to_35                    4
income_household_35_to_50                    4
income_household_50_to_75                    4
income_household_75_to_100                   4
income_household_100_to_150                  4
income_household_150_over                    4
income_household_six_figure                  4
income_individual_median                     1
home_ownership                               4
housing_units                                1
home_value                                   4
rent_median                                  4
rent_burden                                  4
education_less_highschool                    1
education_highschool                         1
education_some_college                       1
education_bachelors                          1
education_graduate                           1
education_college_or_above                   1
education_stem_degree                        1
labor_force_participation                    1
unemployment_rate                            1
self_employed                                4
farmer                                       4
race_white                                   1
race_black                                   1
race_asian                                   1
race_native                                  1
race_pacific                                 1
race_other                                   1
race_multiple                                1
hispanic                                     1
disabled                                     1
poverty                                      4
limited_english                              4
commute_time                                 1
health_uninsured                             1
veteran                                      1
Ozone                                       29
PM25                                        29
N02                                         29
DiagPeriodL90D                               0
dtype: int64


# print the list of numerical columns
numerical_columns = df.select_dtypes(include=['int64', 'float64']).columns
print("List of numerical columns :", numerical_columns)

# print the list of categorical columns
categorical_columns = df.select_dtypes(include=['object']).columns
print("List of categorical columns :", categorical_columns)

List of numerical columns : Index(['patient_id', 'patient_zip3', 'patient_age', 'bmi', 'population',
       'density', 'age_median', 'age_under_10', 'age_10_to_19', 'age_20s',
       'age_30s', 'age_40s', 'age_50s', 'age_60s', 'age_70s', 'age_over_80',
       'male', 'female', 'married', 'divorced', 'never_married', 'widowed',
       'family_size', 'family_dual_income', 'income_household_median',
       'income_household_under_5', 'income_household_5_to_10',
       'income_household_10_to_15', 'income_household_15_to_20',
       'income_household_20_to_25', 'income_household_25_to_35',
       'income_household_35_to_50', 'income_household_50_to_75',
       'income_household_75_to_100', 'income_household_100_to_150',
       'income_household_150_over', 'income_household_six_figure',
       'income_individual_median', 'home_ownership', 'housing_units',
       'home_value', 'rent_median', 'rent_burden', 'education_less_highschool',
       'education_highschool', 'education_some_college', 'education_bachelors',
       'education_graduate', 'education_college_or_above',
       'education_stem_degree', 'labor_force_participation',
       'unemployment_rate', 'self_employed', 'farmer', 'race_white',
       'race_black', 'race_asian', 'race_native', 'race_pacific', 'race_other',
       'race_multiple', 'hispanic', 'disabled', 'poverty', 'limited_english',
       'commute_time', 'health_uninsured', 'veteran', 'Ozone', 'PM25', 'N02',
       'DiagPeriodL90D'],
      dtype='object')
List of categorical columns : Index(['patient_race', 'payer_type', 'patient_state', 'patient_gender',
       'breast_cancer_diagnosis_code', 'breast_cancer_diagnosis_desc',
       'metastatic_cancer_diagnosis_code', 'metastatic_first_novel_treatment',
       'metastatic_first_novel_treatment_type', 'Region', 'Division'],
      dtype='object')


from IPython.display import Image
image_path = "KNN_imputation.png"
#display the image
Image(filename = image_path, width = 800, height = 800)


from sklearn.impute import KNNImputer

# Assuming df is your DataFrame

# Initialize the KNN Imputer
knn_imputer = KNNImputer(n_neighbors=5, weights='uniform', metric='nan_euclidean')

# Use the provided list of numerical columns for KNN imputation
numerical_columns = df.select_dtypes(include=['int64', 'float64']).columns

# Apply KNN imputation
df[numerical_columns] = knn_imputer.fit_transform(df[numerical_columns])


from sklearn.impute import SimpleImputer

# Initialize the Simple Imputer for categorical data
simple_imputer = SimpleImputer(strategy='most_frequent')

# Use the provided list of categorical columns for simple imputation
categorical_columns = df.select_dtypes(include=['object']).columns

# Apply Simple imputation
df[categorical_columns] = simple_imputer.fit_transform(df[categorical_columns])


# Increase display option for maximum rows to ensure all columns are shown
pd.set_option('display.max_rows', 100)

# print list of all 83 columns and null values in each column
print(df.isnull().sum())

patient_id                               0
patient_race                             0
payer_type                               0
patient_state                            0
patient_zip3                             0
patient_age                              0
patient_gender                           0
bmi                                      0
breast_cancer_diagnosis_code             0
breast_cancer_diagnosis_desc             0
metastatic_cancer_diagnosis_code         0
metastatic_first_novel_treatment         0
metastatic_first_novel_treatment_type    0
Region                                   0
Division                                 0
population                               0
density                                  0
age_median                               0
age_under_10                             0
age_10_to_19                             0
age_20s                                  0
age_30s                                  0
age_40s                                  0
age_50s                                  0
age_60s                                  0
age_70s                                  0
age_over_80                              0
male                                     0
female                                   0
married                                  0
divorced                                 0
never_married                            0
widowed                                  0
family_size                              0
family_dual_income                       0
income_household_median                  0
income_household_under_5                 0
income_household_5_to_10                 0
income_household_10_to_15                0
income_household_15_to_20                0
income_household_20_to_25                0
income_household_25_to_35                0
income_household_35_to_50                0
income_household_50_to_75                0
income_household_75_to_100               0
income_household_100_to_150              0
income_household_150_over                0
income_household_six_figure              0
income_individual_median                 0
home_ownership                           0
housing_units                            0
home_value                               0
rent_median                              0
rent_burden                              0
education_less_highschool                0
education_highschool                     0
education_some_college                   0
education_bachelors                      0
education_graduate                       0
education_college_or_above               0
education_stem_degree                    0
labor_force_participation                0
unemployment_rate                        0
self_employed                            0
farmer                                   0
race_white                               0
race_black                               0
race_asian                               0
race_native                              0
race_pacific                             0
race_other                               0
race_multiple                            0
hispanic                                 0
disabled                                 0
poverty                                  0
limited_english                          0
commute_time                             0
health_uninsured                         0
veteran                                  0
Ozone                                    0
PM25                                     0
N02                                      0
DiagPeriodL90D                           0
dtype: int64


# Print descriptive statistics for numerical columns
print("Descriptive Statistics for Numerical Columns:")
print(df[numerical_columns].describe())

Descriptive Statistics for Numerical Columns:
          patient_id  patient_zip3   patient_age           bmi    population  \
count   12906.000000  12906.000000  12906.000000  12906.000000  12906.000000   
mean   547381.196033    573.754300     59.183326     29.033332  20743.600664   
std    260404.959974    275.447534     13.335216      3.819366  13886.694038   
min    100063.000000    101.000000     18.000000     14.000000    635.545455   
25%    321517.000000    331.000000     50.000000     26.684000   9463.896552   
50%    543522.000000    554.000000     59.000000     28.842000  19154.190480   
75%    772671.750000    846.000000     67.000000     31.095500  30021.278690   
max    999896.000000    999.000000     91.000000     85.000000  71374.131580   

            density    age_median  age_under_10  age_10_to_19       age_20s  \
count  12906.000000  12906.000000  12906.000000  12906.000000  12906.000000   
mean    1581.838552     40.502560     11.122744     12.945222     13.290258   
std     2966.217599      4.036952      1.512324      1.923906      3.354000   
min        0.916667     20.600000      0.000000      6.314286      5.925000   
25%      170.876786     37.129825     10.160000     11.741176     11.013415   
50%      700.337500     40.639344     11.039216     12.923944     12.538095   
75%     1666.515385     42.934783     12.190000     14.019767     14.971053   
max    21172.000000     54.570000     17.675000     35.300000     62.100000   

       ...      disabled       poverty  limited_english  commute_time  \
count  ...  12906.000000  12906.000000     12906.000000  12906.000000   
mean   ...     13.335449     13.409128         4.474284     27.978280   
std    ...      3.690846      5.223887         4.836780      5.083756   
min    ...      4.600000      3.433333         0.000000     12.460784   
25%    ...     10.270492      9.663333         0.994444     24.933333   
50%    ...     12.884000     12.177778         2.747222     27.788235   
75%    ...     15.555405     16.642222         5.976000     30.709375   
max    ...     35.155556     38.347826        26.755000     48.020000   

       health_uninsured       veteran         Ozone          PM25  \
count      12906.000000  12906.000000  12906.000000  12906.000000   
mean           8.575330      7.083675     39.826879      7.473876   
std            4.203322      3.109087      3.557957      1.515922   
min            2.440000      1.200000     30.939316      2.636008   
25%            5.618750      4.929688     37.698880      6.642263   
50%            7.465714      6.847059     39.108249      7.686577   
75%           10.617442      8.620000     41.186992      8.276922   
max           27.566102     25.200000     52.237210     11.169408   

                N02  DiagPeriodL90D  
count  12906.000000    12906.000000  
mean      16.092901        0.624516  
std        5.839309        0.484266  
min        2.760371        0.000000  
25%       11.280694        0.000000  
50%       15.529551        1.000000  
75%       20.801880        1.000000  
max       31.504775        1.000000  

[8 rows x 72 columns]


# Calculate the number of rows and columns to fit all numerical columns
n_cols = 4  # Set the number of columns you desire
n_rows = math.ceil(len(numerical_columns) / n_cols)

# Plot histograms
df[numerical_columns].hist(bins=15, figsize=(15, n_rows * 3), layout=(n_rows, n_cols))
plt.tight_layout()
plt.show()


# Ignore specific UserWarnings
warnings.filterwarnings("ignore", message="The figure layout has changed to tight")

plt.figure(figsize=(15, n_rows * 3))  # Adjust figure size based on the number of rows
for i, col in enumerate(numerical_columns):
    plt.subplot(n_rows, n_cols, i + 1)
    sns.boxplot(y=df[col])
    plt.title(col)
    plt.tight_layout()
plt.show()

# Reset warnings to default behavior (optional, if you want to see warnings again after this block)
warnings.filterwarnings("default", message="The figure layout has changed to tight")


pd.set_option('display.max_rows', 100)

# Frequency counts for categorical columns
print("Frequency Counts for Categorical Columns:")
for col in categorical_columns:
    print(f"Column: {col}")
    display(df[col].value_counts())
    print("\n")

Frequency Counts for Categorical Columns:
Column: patient_race

patient_race
White       9973
Black       1056
Hispanic     829
Other        683
Asian        365
Name: count, dtype: int64


Column: payer_type

payer_type
COMMERCIAL            7835
MEDICAID              2569
MEDICARE ADVANTAGE    2502
Name: count, dtype: int64


Column: patient_state

patient_state
CA    2489
TX    1155
NY    1041
MI     858
IL     782
OH     754
FL     609
GA     551
PA     483
MN     377
CO     371
VA     365
IN     317
KY     259
AZ     229
WI     212
WA     200
NC     186
MO     160
NM     141
LA     139
TN     112
SC     110
ID      93
OK      93
OR      89
IA      86
MS      76
MD      71
AR      63
NV      62
KS      50
AL      50
MT      45
NE      40
UT      36
DE      33
WV      27
HI      21
NJ      21
DC      20
SD       8
ND       6
AK       6
WY       4
CT       2
PR       1
RI       1
NH       1
MA       1
Name: count, dtype: int64


Column: patient_gender

patient_gender
F    12906
Name: count, dtype: int64


Column: breast_cancer_diagnosis_code

breast_cancer_diagnosis_code
1749      1982
C50911    1797
C50912    1712
C50919    1467
C50411     978
C50412     877
C50811     491
C50812     419
1744       389
1748       307
C50212     293
C50211     276
C50511     213
C50112     209
C50111     208
C50512     176
C50312     159
C50311     138
C50011     108
C50012     105
1742        98
1741        88
1745        71
C50819      54
C50419      48
C5091       37
1743        34
C50612      29
C50611      25
C50319      16
C50019      13
1746        12
C50219      11
C50119      11
C5041        9
19881        9
C5081        8
C50519       8
C5021        3
C509         3
C50929       3
C50619       3
C5011        2
C50021       1
C5031        1
C5051        1
C50          1
1759         1
C5001        1
C50421       1
Name: count, dtype: int64


# Adjust grid size dynamically based on number of categorical columns
n_cols = 1  # Desired number of columns in the plot
n_rows = (len(categorical_columns) + n_cols - 1) // n_cols  # Calculate required number of rows

plt.figure(figsize=(15, n_rows * 3))

# Suppress warnings for cleaner output
warnings.filterwarnings("ignore", category=UserWarning)

for i, col in enumerate(categorical_columns):
    plt.subplot(n_rows, n_cols, i + 1)
    sns.countplot(y=df[col], order=df[col].value_counts().index)
    plt.title(col)
    plt.tight_layout()

# Show plot
plt.show()

# Reset warnings to default behavior
warnings.filterwarnings("default", category=UserWarning)


import matplotlib.pyplot as plt
import seaborn as sns

# Exclude ID-like columns from correlation analysis if needed
excluded_columns = ['patient_id', 'patient_zip3']
corr_columns = [col for col in numerical_columns if col not in excluded_columns]

plt.figure(figsize=(15, 10))

# Calculate correlation matrix for the relevant numerical columns
corr_matrix = df[corr_columns].corr()

# Generate a heatmap
sns.heatmap(corr_matrix, annot=True, fmt=".2f", cmap='coolwarm')
plt.title("Correlation Matrix")
plt.show()


target = df['DiagPeriodL90D']
# number of unique values in the target column
print(target.unique())

print(target.value_counts())

[1. 0.]
DiagPeriodL90D
1.0    8060
0.0    4846
Name: count, dtype: int64


from sklearn.model_selection import train_test_split
from sklearn.preprocessing import OneHotEncoder, StandardScaler
from sklearn.compose import ColumnTransformer
from sklearn.pipeline import Pipeline

# Define the features and target
X = df.drop('DiagPeriodL90D', axis=1)
y = df['DiagPeriodL90D']

# Split the dataset
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# Preprocessing
numerical_cols = X.select_dtypes(include=['int64', 'float64']).columns
categorical_cols = X.select_dtypes(include=['object']).columns

preprocessor = ColumnTransformer(transformers=[
    ('num', StandardScaler(), numerical_cols),
    ('cat', OneHotEncoder(handle_unknown='ignore'), categorical_cols)
])


from sklearn.linear_model import LogisticRegression
from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier
from sklearn.svm import SVC
from sklearn.neighbors import KNeighborsClassifier
from sklearn.tree import DecisionTreeClassifier
from sklearn.metrics import classification_report

# Define a list of classifiers
classifiers = [
    ('Logistic Regression', LogisticRegression(max_iter=1000)),
    ('Random Forest', RandomForestClassifier()),
    ('SVM', SVC()),
    ('KNN', KNeighborsClassifier()),
    ('Decision Tree', DecisionTreeClassifier()),
    ('nd', GradientBoostingClassifier())
]

# Iterate over the defined classifiers
for name, classifier in classifiers:
    # Create the pipeline with the current classifier
    pipeline = Pipeline([
        ('preprocessor', preprocessor),
        ('classifier', classifier)
    ])
    
    # Train the model using the training data
    pipeline.fit(X_train, y_train)
    
    # Generate predictions on the test data
    y_pred = pipeline.predict(X_test)
    
    # Print the classification report for the current model
    print(f"{name} Performance:\n")
    print(classification_report(y_test, y_pred))
    print("--------------------------------------------------\n")

Logistic Regression Performance:

              precision    recall  f1-score   support

         0.0       0.89      0.58      0.70       965
         1.0       0.79      0.96      0.87      1617

    accuracy                           0.82      2582
   macro avg       0.84      0.77      0.78      2582
weighted avg       0.83      0.82      0.80      2582

--------------------------------------------------

Random Forest Performance:

              precision    recall  f1-score   support

         0.0       0.80      0.58      0.67       965
         1.0       0.78      0.91      0.84      1617

    accuracy                           0.79      2582
   macro avg       0.79      0.74      0.76      2582
weighted avg       0.79      0.79      0.78      2582

--------------------------------------------------

SVM Performance:

              precision    recall  f1-score   support

         0.0       0.90      0.54      0.67       965
         1.0       0.78      0.96      0.86      1617

    accuracy                           0.80      2582
   macro avg       0.84      0.75      0.77      2582
weighted avg       0.82      0.80      0.79      2582

--------------------------------------------------

KNN Performance:

              precision    recall  f1-score   support

         0.0       0.52      0.41      0.46       965
         1.0       0.69      0.77      0.73      1617

    accuracy                           0.64      2582
   macro avg       0.60      0.59      0.59      2582
weighted avg       0.62      0.64      0.63      2582

--------------------------------------------------

Decision Tree Performance:

              precision    recall  f1-score   support

         0.0       0.60      0.63      0.61       965
         1.0       0.77      0.75      0.76      1617

    accuracy                           0.70      2582
   macro avg       0.68      0.69      0.69      2582
weighted avg       0.71      0.70      0.70      2582

--------------------------------------------------

Gradient Boosting Performance:

              precision    recall  f1-score   support

         0.0       0.90      0.57      0.70       965
         1.0       0.79      0.96      0.87      1617

    accuracy                           0.82      2582
   macro avg       0.84      0.77      0.78      2582
weighted avg       0.83      0.82      0.80      2582

--------------------------------------------------


import pandas as pd

# Initialize an empty list to store results
results = []

# Iterate over classifiers
for name, classifier in classifiers:
    pipeline = Pipeline([
        ('preprocessor', preprocessor),
        ('classifier', classifier)
    ])
    
    pipeline.fit(X_train, y_train)
    y_pred = pipeline.predict(X_test)
    
    # Generate classification report as a dictionary
    report = classification_report(y_test, y_pred, output_dict=True)
    
    # Extract relevant metrics
    precision_0 = report['0.0']['precision']
    recall_0 = report['0.0']['recall']
    f1_score_0 = report['0.0']['f1-score']
    
    precision_1 = report['1.0']['precision']
    recall_1 = report['1.0']['recall']
    f1_score_1 = report['1.0']['f1-score']
    
    accuracy = report['accuracy']
    macro_avg_precision = report['macro avg']['precision']
    macro_avg_recall = report['macro avg']['recall']
    macro_avg_f1 = report['macro avg']['f1-score']
    
    weighted_avg_precision = report['weighted avg']['precision']
    weighted_avg_recall = report['weighted avg']['recall']
    weighted_avg_f1 = report['weighted avg']['f1-score']
    
    # Append results to the list
    results.append([name, precision_0, recall_0, f1_score_0, precision_1, recall_1, f1_score_1,
                    accuracy, macro_avg_precision, macro_avg_recall, macro_avg_f1,
                    weighted_avg_precision, weighted_avg_recall, weighted_avg_f1])

# Create a DataFrame from the results
results_df = pd.DataFrame(results, columns=['Model', 'Precision (Class 0)', 'Recall (Class 0)', 'F1-Score (Class 0)',
                                            'Precision (Class 1)', 'Recall (Class 1)', 'F1-Score (Class 1)',
                                            'Accuracy', 'Macro Avg Precision', 'Macro Avg Recall', 'Macro Avg F1',
                                            'Weighted Avg Precision', 'Weighted Avg Recall', 'Weighted Avg F1'])

# Display the results DataFrame
display(results_df)


# display transpose of the results_df dataframe
display(results_df.T)


import matplotlib.pyplot as plt

# Set the style
plt.style.use('seaborn-darkgrid')

# Create a color palette
palette = plt.get_cmap('Set1')

# Plot each column
num_metrics = results_df.shape[1] - 1  # Exclude the 'Model' column
fig, axes = plt.subplots(num_metrics, 1, figsize=(10, 5 * num_metrics))

metrics = results_df.columns[1:]  # Exclude 'Model' column

for i, metric in enumerate(metrics):
    for j in range(results_df.shape[0]):
        axes[i].plot(results_df['Model'], results_df[metric], marker='', color=palette(j), linewidth=2.5, alpha=0.9, label=metric)
    axes[i].set_title(metric)
    axes[i].set_xticks(results_df['Model'])
    axes[i].set_xticklabels(results_df['Model'], rotation=45, ha="right")
    axes[i].legend(loc='upper left', ncol=2)

# Show plot
plt.tight_layout()
plt.show()


from IPython.display import Image
image_path = "Kaggle_leaderboard.png"
#display the image
Image(filename = image_path, width = 1200, height = 800)

	patient_id	patient_race	payer_type	patient_state	patient_zip3	patient_age	patient_gender	bmi	breast_cancer_diagnosis_code	breast_cancer_diagnosis_desc	...	disabled	poverty	limited_english	commute_time	health_uninsured	veteran	Ozone	PM25	N02	DiagPeriodL90D
0	475714	NaN	MEDICAID	CA	924	84	F	NaN	C50919	Malignant neoplasm of unsp site of unspecified...	...	12.871429	22.542857	10.100000	27.814286	11.200000	3.500000	52.237210	8.650555	18.606528	1
1	349367	White	COMMERCIAL	CA	928	62	F	28.49	C50411	Malig neoplm of upper-outer quadrant of right ...	...	8.957576	10.109091	8.057576	30.606061	7.018182	4.103030	42.301121	8.487175	20.113179	1
2	138632	White	COMMERCIAL	TX	760	43	F	38.09	C50112	Malignant neoplasm of central portion of left ...	...	11.253333	9.663333	3.356667	31.394915	15.066667	7.446667	40.108207	7.642753	14.839351	1
3	617843	White	COMMERCIAL	CA	926	45	F	NaN	C50212	Malig neoplasm of upper-inner quadrant of left...	...	8.845238	8.688095	5.280952	27.561905	4.404762	4.809524	42.070075	7.229393	15.894123	0
4	817482	NaN	COMMERCIAL	ID	836	55	F	NaN	1749	Malignant neoplasm of breast (female), unspeci...	...	15.276000	11.224000	1.946000	26.170213	12.088000	13.106000	41.356058	4.110749	11.722197	0

	Model	Precision (Class 0)	Recall (Class 0)	F1-Score (Class 0)	Precision (Class 1)	Recall (Class 1)	F1-Score (Class 1)	Accuracy	Macro Avg Precision	Macro Avg Recall	Macro Avg F1	Weighted Avg Precision	Weighted Avg Recall	Weighted Avg F1
0	Logistic Regression	0.887302	0.579275	0.700940	0.792008	0.956092	0.866349	0.815259	0.839655	0.767683	0.783645	0.827623	0.815259	0.804529
1	Random Forest	0.801724	0.578238	0.671884	0.784199	0.914657	0.844419	0.788923	0.792962	0.746448	0.758152	0.790749	0.788923	0.779936
2	SVM	0.897569	0.535751	0.670993	0.776670	0.963513	0.860061	0.803641	0.837120	0.749632	0.765527	0.821855	0.803641	0.789398
3	KNN	0.519104	0.408290	0.457077	0.686780	0.774273	0.727907	0.637490	0.602942	0.591282	0.592492	0.624113	0.637490	0.626686
4	Decision Tree	0.597793	0.617617	0.607543	0.767192	0.752010	0.759525	0.701782	0.682493	0.684813	0.683534	0.703881	0.701782	0.702723
5	Gradient Boosting	0.898858	0.570984	0.698352	0.789741	0.961657	0.867262	0.815647	0.844300	0.766321	0.782807	0.830523	0.815647	0.804133

	0	1	2	3	4	5
Model	Logistic Regression	Random Forest	SVM	KNN	Decision Tree	Gradient Boosting
Precision (Class 0)	0.887302	0.801724	0.897569	0.519104	0.597793	0.898858
Recall (Class 0)	0.579275	0.578238	0.535751	0.40829	0.617617	0.570984
F1-Score (Class 0)	0.70094	0.671884	0.670993	0.457077	0.607543	0.698352
Precision (Class 1)	0.792008	0.784199	0.77667	0.68678	0.767192	0.789741
Recall (Class 1)	0.956092	0.914657	0.963513	0.774273	0.75201	0.961657
F1-Score (Class 1)	0.866349	0.844419	0.860061	0.727907	0.759525	0.867262
Accuracy	0.815259	0.788923	0.803641	0.63749	0.701782	0.815647
Macro Avg Precision	0.839655	0.792962	0.83712	0.602942	0.682493	0.8443
Macro Avg Recall	0.767683	0.746448	0.749632	0.591282	0.684813	0.766321
Macro Avg F1	0.783645	0.758152	0.765527	0.592492	0.683534	0.782807
Weighted Avg Precision	0.827623	0.790749	0.821855	0.624113	0.703881	0.830523
Weighted Avg Recall	0.815259	0.788923	0.803641	0.63749	0.701782	0.815647
Weighted Avg F1	0.804529	0.779936	0.789398	0.626686	0.702723	0.804133

reference: blog.dailydoseofds.com¶

KNN Imputation for Numerical Columns¶

Simple Imputation for Categorical Columns¶

EDA¶

Interpretations¶

Discussions¶

Histograms for Numerical Columns¶

Box Plots for Numerical Columns¶

Interpretations¶

Discussions¶

Bar Charts for Categorical Columns¶

Correlation Matrix Visualization¶

ML model training and Evaluation¶

Step 1: Data Preparation¶

Step 2: Implement Classification Models¶

Interpretations¶

Discussions¶